We'll respond shortly.
Many people new to building apps fall in love the moment they learn about the idea of a Minimum Viable Product. “It’s minimal! So there’s less risk. And it’s viable! So it’ll prove something!”. Unfortunately, it’s easy for the line of “minimum” or “viable” to slip. How can a team stay focused?
Lean Hypothesis are an effective way to help the team connect the problem they’re trying to solve to the product they’re building. They take the form of
"We believe [TYPE OF USER] has a problem [DOING THING]. We can help them with [OUR SOLUTION]. We'll know we're right if [CHANGE IN METRIC]."
Let’s say we’ve got a problem on Hamazon, our e-commerce platform. Users are clicking around and spending lots of time-on-site, but they’re not putting items in their cart and converting. We might start with
"We believe SHOPPERS have a problem CHOOSING FROM THE MASSIVE SELECTION OF PORK PRODUCTS. We can help them by SHOWING RECOMMENDATIONS FROM OTHER SHOPPERS. We'll know we're right if SHOPPERS ADD ITEMS TO CART MORE QUICKLY".
Once we have a hypothesis, an MVP can be defined as the least amount of work we can do to in/validate the hypothesis. We started from the assumption that we need a recommendation engine, but rather than building it out (an expensive proposition), we’ve honed in on the a more specific problem: shoppers need guidance. For much less effort, we could test this hypothesis by curating and featuring a small selection of recommended products—this “Recommended Products” section is our new MVP! This might fail to impel shoppers to add items to their cart more quickly; if so, we’ll try a different hypothesis. (Notice we didn’t incur the cost of building a recommendation engine!). But if it succeeds, we’ll be able to iterate. Maybe we find that the Recommended Products converts very well for users from the east coast, but not so well for users from the west coast.
"We believe WEST COAST SHOPPERS have a problem WITH THE RECOMMENDED PRODUCTS WE'RE CHOOSING. We can help them by GIVING DIFFERENT RECOMMENDATIONS. We'll know we're right if WEST-COAST SHOPPERS ADD ITEMS TO CART MORE QUICKLY".
So now our MVP would be a feature that lets our in-house curation team target separate Recommendation sets based on geography. We’d continue to iterate, and it’s possible our recommendation targeting would get so specific that we’d end up building a recommendation engine, but we’d only do so if the business needs led us there, rather than our intuition. In that way, we’d iterate towards a truly minimal, truly viable product.